Early risk stratification for ventilator-associated pneumonia in mechanically ventilated neonates: development and external validation of a multicenter prediction model

Ventilator-associated pneumonia (VAP) is an important healthcare-associated infection in mechanically ventilated neonates. Early identification of high-risk neonates may help guide targeted prevention in neonatal intensive care units. This study aimed to develop and externally validate a prediction model for VAP using routinely available early clinical and laboratory variables. This multicenter retrospective study used data from five hospitals in China: four for model development and one independent hospital for external validation. Eligible neonates remained invasively ventilated for more than 48 h, and risk was estimated at 48 h after intubation using clinical variables available before intubation and the first laboratory measurements obtained within 24 h. Univariable logistic regression and least absolute shrinkage and selection operator regression were used for predictor screening, followed by multivariable logistic regression. Discrimination, calibration and clinical utility were assessed using receiver operating characteristic curves, calibration curves, the Hosmer–Lemeshow test, decision curve analysis and clinical impact curves. Bootstrap bias-corrected calibration curves were obtained using 1000 resamples in both cohorts. The study was retrospectively registered in the Medical Research Registration Information System (registration no. MR-50-26-037954). A total of 669 neonates were included: 363 in the development cohort and 306 in the external validation cohort. VAP occurred in 124 and 100 neonates, respectively. Five predictors were retained: white blood cell count, hemoglobin, albumin, C-reactive protein and pre-intubation sepsis. The model showed good discrimination, with an area under the curve of 0.854 (95% confidence interval, 0.813–0.895) in the development cohort and 0.821 (95% confidence interval, 0.766–0.876) in the external validation cohort. Calibration was acceptable in both cohorts. At the fixed cutoff of 0.329, sensitivity and specificity were 69.0% and 85.4%, respectively, in the external validation cohort, and VAP occurred in 69.7% of high-risk neonates compared with 15.0% of low-risk neonates. A five-variable model showed good discrimination and acceptable calibration for early VAP risk stratification and may support risk-informed surveillance and reinforcement of preventive practices; prospective impact evaluation is required before routine implementation.

Authors

Institutions

Publication Details

Journal
Antimicrobial Resistance and Infection Control
Published
2026-11-16
DOI
https://doi.org/10.1186/s13756-026-01863-z
Primary Topic
Nosocomial Infections in ICU
Type
article
Field-Weighted Citation Impact
0.00
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
article

Early risk stratification for ventilator-associated pneumonia in mechanically ventilated neonates: development and external validation of a multicenter prediction model

Ruixue Zheng, Min Tao, Shiling Yan, Chenghuan Zhang et al.
3 links
Antimicrobial Resistance and Infection Control
Nosocomial Infections in ICU
article

Early risk stratification for ventilator-associated pneumonia in mechanically ventilated neonates: development and external validation of a multicenter prediction model

Ruixue Zheng, Min Tao, Shiling Yan, Chenghuan Zhang, Qingling Xu, Liu Wang, Leilei Shen
article en

Abstract

Ventilator-associated pneumonia (VAP) is an important healthcare-associated infection in mechanically ventilated neonates. Early identification of high-risk neonates may help guide targeted prevention in neonatal intensive care units. This study aimed to develop and externally validate a prediction model for VAP using routinely available early clinical and laboratory variables. This multicenter retrospective study used data from five hospitals in China: four for model development and one independent hospital for external validation. Eligible neonates remained invasively ventilated for more than 48 h, and risk was estimated at 48 h after intubation using clinical variables available before intubation and the first laboratory measurements obtained within 24 h. Univariable logistic regression and least absolute shrinkage and selection operator regression were used for predictor screening, followed by multivariable logistic regression. Discrimination, calibration and clinical utility were assessed using receiver operating characteristic curves, calibration curves, the Hosmer–Lemeshow test, decision curve analysis and clinical impact curves. Bootstrap bias-corrected calibration curves were obtained using 1000 resamples in both cohorts. The study was retrospectively registered in the Medical Research Registration Information System (registration no. MR-50-26-037954). A total of 669 neonates were included: 363 in the development cohort and 306 in the external validation cohort. VAP occurred in 124 and 100 neonates, respectively. Five predictors were retained: white blood cell count, hemoglobin, albumin, C-reactive protein and pre-intubation sepsis. The model showed good discrimination, with an area under the curve of 0.854 (95% confidence interval, 0.813–0.895) in the development cohort and 0.821 (95% confidence interval, 0.766–0.876) in the external validation cohort. Calibration was acceptable in both cohorts. At the fixed cutoff of 0.329, sensitivity and specificity were 69.0% and 85.4%, respectively, in the external validation cohort, and VAP occurred in 69.7% of high-risk neonates compared with 15.0% of low-risk neonates. A five-variable model showed good discrimination and acceptable calibration for early VAP risk stratification and may support risk-informed surveillance and reinforcement of preventive practices; prospective impact evaluation is required before routine implementation.

Antimicrobial Resistance and Infection Control
Army Medical University (CN), Chongqing Maternal and Child Health Hospital (CN), Chinese People's Liberation Army (CN), Chongqing Medical University (CN)
Good health and well-being
Nosocomial Infections in ICU
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Laika access coming soon